Diauxic shift of Roseburia intestinalis

Enables slow growth after depletion of glucose

Introduction to Subject

R. intestinalis survives post-glucose (and post-mucin-degradation products) depletion by entering a slow-growth mode.

A combination of targeted metabolomics and RNA-seq showed that the slow growth mode of R. intestinalis represents a diauxic shift towards acetate and lactate consumption (both leading to butyrate production).

R. intestinalis consistently increased the expression of genes involved in glycolysis, pyruvate utilization, mucin hydrolysis, galactose and mannose utilization, and the butyrate-producing process after depletion of glucose/pyruvate.

Liu, B., Garza, D.R., Gonze, D. et al. Starvation responses impact interaction dynamics of human gut bacteria Bacteroides thetaiotaomicron and Roseburia intestinalis. ISME J 17, 1940–1952 (2023). https://doi.org/10.1038/s41396-023-01501-1

Data overview

  • Fermentation

    • containing 888 experiments with data:
    • Exp, Replicate, Mucin, Species.combi, T0h, T4h, T6h, T8h, T10h, T12h, T14h, T16h, T20h, T24h, T28h, T32h, T36h, T38h, T40h, T42h, T44h, T48h, T56h, T60h, T72h, T86h, T96h, T120h, T144h, T156h, T168h, T192h, T216h, T240h, T264h, T288h

  • RNA expression data

    • containing recorded expression of 4262 genes in 45 experiments

  • Annotation for expression dataset

    • Coupling internal gene ID’s to gene names , KO ID’s, and COG ID’s

  • Meta-data for fermentation dataset

    • Coupling RNA expression samples with fermentation batches and time points

Data Cleaning and Augmentation

Removing empty time-points from fermentation dataset

ri_fermentation_data |>
  filter(Exp == "RI_batch_WC1") |>
  select( "Rowname", (starts_with("T") & ends_with("h")) ) |>
  select(where(~!all(is.na(.x)) ) ) |>
  write_csv("../data/02_fermentation_data_clean.csv")

Subsetting RNA expression to wild type R. intestinalis only

ri_expression_data |> 
  pivot_longer(cols = -X,
               
               names_to = "sample",
               values_to = "expression") |> 
  inner_join(ri_metadata, by = c("sample" = "GC.code" )) |> 
  filter(Exp == "RI_batch_WC") |> 
  select(X,expression,sample) |> 
  pivot_wider(names_from = sample,
              values_from = expression) |> 
  write_csv("../data/02_data_clean.csv")

Correlation Matrix from Gene Expression data

  • Input data was a total of 9 samples across 3 time conditions
  • Matrix reveals co-expression patterns in R. intestinalis gene expression data
  • Co-regulated genes can potentially contribute to the same pathways

K-Means Identifies 3 Distinct Clusters

  • Methodology: Applied K-means clustering to the correlation matrix.
  • Observation: Genes show clear separation in 2D PCA space.

Note: The cutoff for 3 clusters was selected for regulation validation purposes.

Figure 1: PCA Projection of Gene Clusters

Pathway Enrichment Reveals Distinct Cluster Functions

Observations

  • Major pathways (top bars) are shared across all three clusters.
  • Certain pathways are exclusive to Cluster 1 or 2 (e.g., lower yellow/green bars).

Expression

Top Counts

Fermentation

R. intestinalis’ active cell count remains stable, even after the nutrients are depleted. This showcases this species’ resilience. Butyrate concentration grows as glucose levels are dropping, which is expected, as butyrate is the fermentation product of R. intestinalis. The exact mechanism of its carbon uptake after glucose levels have dropped remains elusive

Conclusion